Colleges Combat AI Cheating: Blue Books, Flawed Detection Software, and False Positives

The Rise of AI in Academia

• 🤖 Generative AI tools like ChatGPT are revolutionizing writing, capable of creating essays and code in seconds.
• 🎓 A survey found that 90% of college students have used AI for homework, and nearly half have used it during tests.

Tackling AI: Traditional and Technological Approaches

• 📚 Sales of "blue books" (lined composition notebooks for handwritten exams) are increasing as a tactic to deter AI use.
• 🔍 Schools are employing AI detection software to maintain academic integrity, but these tools are not always accurate.

The Problem of False Positives

• ⚠️ False positives occur when AI detection software incorrectly flags original student work as AI-generated.
• 📉 One student was wrongly accused of plagiarism by AI detection software, risking his scholarship and academic standing.
• 📊 Even experts like Dr. Sohel Fazy found that popular AI detection tools labeled his own text as AI-generated, highlighting the unreliability of these systems.
• 🏢 Companies like Turn It In claim low false positive rates (e.g., 1%), but educators caution against relying solely on these scores.

Challenges and Evolving Strategies

• 💬 Students with English as a second language are disproportionately flagged by AI detectors due to natural variations in sentence construction and word usage.
• 🔒 Some institutions, like John J. College, are cautioning faculty against over-reliance on AI detection tools.
• ✍️ Strategies like using Google Docs to track edits word-by-word are being used to prove authenticity.
• 🔄 The rapid evolution of AI technology makes it a constant game of defense for academic institutions seeking to protect integrity.